MOON – Meteorological Services Workshop
Working-Group 1 : Requirements for Surface Meteorological Data
- Nadia Pinardi
- Pierre Blouch
- Fabrice Lecornu
- Sergio Pasquini
- Roberto Bozzano
- Rosalia Santoleri
- Leonidas Perivoliotis
- George Zodiatis
- Paolo Pagano
I - Basic state variables needed
I-1. From atmospheric models
Analysis and forecasts from LAM (forcing for the oceanic models), Analysis and
Forecasts from ECMWF. Coherency required between LAM and global models.
The variables have not been sorted by order of priority:
- Surface winds (10 metres is the standard height)
- Dew point temperature and relative humidity (10 metres should be suitable)
- Short wave radiation net and downward
For biochemical forecasts, some wave lentghs are more important than
others. Ask NWP centres if they may provide a discrete spectra (to be
checked). If available PAR is wishable.
- Long wave net and downward
- Total cloud coverage + height of the base of the lowest ones
- Total Precipitation
- Total water vapour content (to correct satellite SST)
- Turbulent heat fluxes(see §3). MOON partners wish to re-compute them
themselves (bulk functions), but this provision could help for verifications
- Wind stress (same context than above : for verification after re-computation)
- Waves (height, direction, period and spectra)
- Wind drag coefficient
Time resolution : 1 hour for analysis and forecasts (at least for short term forecasts) –
instead of 3h or 6h presently.
Limited domain: atmospheric boundary layer, Mediterranean Sea, short time
Land-sea masks and interpretation tools for each model used.
I-2. From in situ observing stations (meteo-oceanographic measurements)
Requirement: All measured variables sent onto the GTS.
P. Blouch inform the workgroup that wind data reported onto the GTS are those
measured at the level of the anemometers, without any correction for their height
(WMO recommendation). WMO (for VOS) and JCOMMOPS (for buoys) normally
maintain stations metadata which contain the heights of sensors. In future formats
(BUFR, mandatory in force after 31st December xxxx), should contain the metadata
associated to the data.
In situ SST data are strongly required to validate satellite SST – or correct it from the
atmospheric attenuation due to water vapour or dust. Moored and drifting buoys SST
data are more valuable than ship’s SST. Metadata exists for these latest, giving the
method and the depth of measurement.
- Which height should the state variables be taken?
Requirement from MOON : Parameters should be provided at the same level (e.g.
10 metres). However, it appears that meteorological standard outputs are 2
metres for air temperature and humidity. Pressure is at sea level, wind is at 10
metres. How to retrieve air temperature and humidity at 10 metres if provided at 2
metres ? Would it be possible to directly get them at 10 metres ? Discussion
continued on wishes to have a full data set in the atmospheric boundary layer or a
way to deduce them from a partial set.
- Which correction should be used for clouds and precipitation data? Feedback from
Met Services to MOON with regards how to use forecasted clouds and
precipitation values. Assessment of their quality. Verifications thanks to
observations including submarine acoustic technique.
- Precipitation is required under different forms:
- satellite observations
- model analysis and forecasts
- Which Albedo data should be used? How the LAM and ECMWF use defined albedo
of the sea.
- River run off
Presentation of HYPROM (NMM + hydrological model) by Nadia. Cooperation
with Serbia. Excellent results. F. Lecornu provides the workgroup with PreviMer
experience (Cemagref GR4J). Such models have been calibrated on small rivers
in Brittany. They are difficult to use on large river basins. Requirement: ask Met
Services and hydrological agencies to provide MOON with river run off forecasts.
Example: SIM model from SCHAPI, France. HyMEX is a relevant framework for
- Do atmospheric models forecast the SST ?
E-SURFMAR asked NWP centres how their SST analysis are made and, in
particular, if they are assimilating SST observations reported by ships. Replies are
given in a table. The experts who were questionned will be asked if their SST are
forecasted as for other parameters.
- Which parameters for satellite SLA (Sea Level Anomalies) ?
(to be completed)
II - Exchange protocols for atmospheric model forcing data
Recommendations from MOON
- Timely release (one hour) after atmospheric forecast alone is done
- Format: GRIB
- FTP protocol
- Data policy
MOON DEA (Data Exchange Agreement) should be used by Met Offices. Follow
the GMES development in terms of data policy. Specific agreement between
ECMWF and MOON should be considered.
III - Exchange protocols for met-oceanographic observations data
III-1. From observing systems to meteorological services
Recommendation to MOON partners to report the data of their platforms in real-time
onto the GTS (Global Telecommunication System of WMO). Done for Puertos del
Estado moored buoy network and Italia-1. To be done for Poseidon network, Venice
platform and other Italian buoys. This should be done in BUFR format through the
nearest National Meteorological Service (NMS). Arrangement should be found
between the basic data provider and the NMS to code the data.
WMO alphanumeric codes (SHIP, BUOY, TESAC (e.g. for XBTs and Argo floats),
TRACKOB (e.g. thermosalinographs), WAVEOB (for wave spectra data) should
disappear within a couple of years. It is recommended to use BUFR as matter of
priority. Not necessary to develop in old formats. The BUFR template for buoys must
be used to report wave spectra data (exclusively used by Spain).
The GTS is the core of the future WMO Information System (WIS). The WIS should
become the main transmission support for weather, climate, water (including ocean),
environmental data and natural disaster information in the frame of GEOSS.
Check whether observation data sent onto the GTS by MOON are essential
according to WMO resolution 40.
Data policy: decided in MOON DEA
Timely release (depending on the parameters – to be developed
Format: althought NetCDF would be desirable for MOON (SeaDataNet standard), the
WMO standard observation format is BUFR (as GRIB for model data).
Recommendation: Universal BUFR NetCDF converters should be searched or
III-2. From meteorological services to MOON
Find the proper way for meteorological services to distribute surface marine and
oceanographic observation data running onto the GTS (then the WIS) to the MOON
partners. In return of their observation data sent onto the GTS, partners may have
access to the whole met-oceanographic data required for the Mediterranean Sea.
Data users should address their requirements to the NMS of their country which
should give them access to the data… (to be checked with Eumetnet members).
Data policy: should be in accordance with WMO Resolution 40.
III-3. Oceanographic model data
One can imagine the transmission of oceanographic model data on the WIS later on.
IV – Observing system requirements
- SOOP and VOS system to be re-assessed and integrated with Ferry Box for
maximum benefit (i.e. instruments installed on a same ship).
E-SURFMAR proposes to procure Shipborne Automated Weather Sations (S-
AWS) to equip ships plying in the Mediterranean Sea:
- A BaTos AWS - integrated station measuring pressure, air temperature
and humidity, wind and possibly SST and allowing visual parameters
added before transmission - could be procured to CNR for RV Urania
(status of the project to be checked).
- Fifteen BaRos AWS (autonomous station measuring air pressure only)
are available for ships in the Mediterranean Sea. They only require
power supply from the ships. May be considered as a starting initiative.
Easy to install, they could be replaced by more complete S-AWS in the
future: possibility to connect other sensors than a barometer: wind, air
temperature and humidity, radiations...
These S-AWS should be installed by MOON partners. E-SURFMAR will
support the communication costs.
MOON assembly in March 2010, to approve a plan to be submitted to
Eumetnet. Eumetnet Programme Board for Observation (PB-Obs) meet on 22-
24th of March 2010.
- Moored Buoy to be coordinated at the basin scale level.
Moored buoy systems in the Mediterranean
MOON has developed in the past ten years real time Multisensor Moored
Array (M3A) stations which are transmitting in real time meteo-oceanographic
measurements (part of the EuroSites Network too)
MOON Members are responsible for national coastal buoy networks which
collect meteo-oceanographic data in real time and make it available with the
MOON-DEA to MOON Members
It is recommended that all the MOON and national Mediterranean real time
buoy data should be delivered in BUFR format to the corresponding
Meteorological Offices to be transmitted through the GTS
In addition it is recommended that the GTS meteo-oceanographic Information
is made available through each relevant national meteorological office to
- Satellite SST to be exchanged in real time from MyOcean TAC
There is a lack of in situ SST measurements in the Mediterranean Sea to
calibrate and validate SST satellite data. MSG Seviri is the most recent
satellite SST sensor. SST fields produced by MyOcean should be available to
Foreseen works: assessing the observation of the diurnal cycle of SST
through satellite measurements.
MOON has developed the MyOcean service for SST at the basin scale level
MSG SEVIRI data is used in the multi-sensor satellite L4 product from MyOcean
These satellite products should be better used by the Mediterranean atmospheric
NWP community through the MyOcean central desk
Several improvements should be encouraged in the future:
Modelling of the daily cycle by usage of GTS SST and SEVIRI data
Higher resolution products (1 km)
- Wave data to be exchanged for wave model validation (ESEOO system, ISPRA
system, POSEIDON) (see § III-1.)
- Scatterometer winds (to be discussed next time)
- Radiation measurements: they could be done on ships equipped with an S-AWS
station funded by E-SURFMAR (to be explored).
- Cost effectiveness of observation data transmission. Compression method must be
applied if not yet done. Iridium appears a suitable cost effective
communication system. The volume of data is related to the measurements,
and the resolutions. Large differences between the different stations exist.
- Connection with MyOcean in situ TAC for the Meditteranean should be explored in
Technology developments (for information)
- Sea temperature sampled more frequently than through Argo floats (e.g. every hour
instead of every 5 days), over the first 80 metres for instance. This would allow
to observe the diurnal cycle as well as wind forced events. Drifting buoys may
efficiently do that at a reasonnable cost.
- For information : existence of salinity drifting buoys which will be used to calibrate
and validate SMOS. Some will be deployed for HyMEX.
- XBT with temperature and fluorimetry (T/Flap)
Other requirements to Met Services
Sand and dust transports from models and satellite data.